Custom reluctor ring, suggestions on where to get one

-

DionR

Well-Known Member
Joined
May 7, 2008
Messages
4,831
Reaction score
2,358
Location
Spokane, WA
Need a custom ring made to move the Magnum crank pickup to the front of my motor. It's not hard to draw one up and get it cut in a water jet, but I thought there was a place that would make a custom ring and save me some trouble.

Can't seem to find them now.

Anyone got a suggestion on where to get one made?
 
It's not hard to draw one up and get it cut in a water jet ....
It is harder than that. The diameter must shrink, and it is not possible to just scale it down. The sensor diameter influences the trigger points, so the sensor output does not exactly track dimensions of openings and tabs. The electrical signal duration shrinks for openings, and extends for tabs. The ECU, is very application specific, so signals need to be very close to OEM, or there will be problems. Yes it is possible, you just have to understand sensing methods, and make adjustments, in ring design and or sensor choice.
 
Oh hell, Dave you had to toss a wrench in the "wheel" din'ja? LOL I did not know that. I assumed if it had so many triger points, it would be "good." I guess in the Holley / Megasquirt / software, you can change those settings
 
It is harder than that. The diameter must shrink, and it is not possible to just scale it down. The sensor diameter influences the trigger points, so the sensor output does not exactly track dimensions of openings and tabs. The electrical signal duration shrinks for openings, and extends for tabs. The ECU, is very application specific, so signals need to be very close to OEM, or there will be problems. Yes it is possible, you just have to understand sensing methods, and make adjustments, in ring design and or sensor choice.

Huh? Why would it care if it was 5 degrees of rotation on a 12 inch wheel versus a 12 foot wheel? As long as it was the same number of degrees, how would it know? It's late so maybe I am missing something, but if they are both rotating at the same rpm, there wouldn't be any difference between the rise and fall of the signal, right?

Not trying to be argumentative, I know you have far more experience in this, just not seeing how the ECM would even know.

Here is what I know about the stock Magnum system. The flex plate I measured was 11.75" in diameter and the notches where .4" wide. This calculated out to almost exactly 4 degrees of rotation. I also know that a manual trans motor uses a different pickup than an automatic trans motor, and the difference is in the length of the sensor, suggesting that the diameter of the ring on a flywheel is different than that of a flex plate. Don't have data on the flywheel so I don't know how it compares.

Not certain, but seems like I have been told that the Magnum ECM only cares about one edge of the notch.

Also, I am aware that the notch can shrink to where the computer can't see it, so I am being careful to avoid that.
 
I think you completely missed my point. Yes the edges are important, we agree on that. What happens is that the sensor does not trigger at the center of the sensor, but before, and it releases past center.
It is not fractions of a degree, but it is few to several degrees.

The reluctor should be low carbon anealed steel. Runout, thickness are important too, because they impact operate points.

Sensors have data sheets, that help with applications, but most are not readily available for automotive parts. You can learn from reading sheets for similar sensors used in automation industry.

A common method is to make a wheel of correct diameter, thickness, shape ... setup sensor, capture signals with digital scope or logic analyzer. From that, make adjustments in tab, and opening widths. Then verify.

Starting point in your case is knowing the precise OEM signals.
 
I think you completely missed my point. Yes the edges are important, we agree on that. What happens is that the sensor does not trigger at the center of the sensor, but before, and it releases past center.
It is not fractions of a degree, but it is few to several degrees.

This is a Hall Effect sensor, I thought those were pretty much instantaneous?

So what you are saying is that it's the distance past the edges that is the issue, and the diameter of the wheel won't change that distance, even though the number of degrees it rotates will change to get the distance.

Starting point in your case is knowing the precise OEM signals.

Here is what I got from a Mopar engineer years ago in this regards:

"The good news is that it's much simpler than you are thinking. The JTEC is counting on/off pulses. It doesn't care about the width. If you carve 8 equally spaced notches in the pulley/tone ring, then the JTEC will be happy.
Try and line up your new crank trigger to a notch the same time the OEM crank trigger is on a notch. You should have good luck. You will need to adjust the timing by trial and error, but a trip to the drag strip will find the best setting."
 
I can get one done. I sell these 36-1 wheels here: [FOR SALE] - 36-1 Crank Trigger Wheels 318/340/360/383/440

I believe the factory Magnum reluctor had 8 holes, but I'm not sure of the overall diameter of the ring? Size of the holes and overall diameter would come in to play when designing a smaller one to be pressed on to the crank pulley.

Great! I need to get an accurate measurement of the pulley OD because I know it is bigger than 4.26". The OD I need is 7.75", and it needs 8 equally spaced notches, but I haven't worked out what size yet. Will be in contact.

Thanks!
 
This is a Hall Effect sensor, I thought those were pretty much instantaneous?

So what you are saying is that it's the distance past the edges that is the issue, and the diameter of the wheel won't change that distance, even though the number of degrees it rotates will change to get the distance.



Here is what I got from a Mopar engineer years ago in this regards:

"The good news is that it's much simpler than you are thinking. The JTEC is counting on/off pulses. It doesn't care about the width. If you carve 8 equally spaced notches in the pulley/tone ring, then the JTEC will be happy.
Try and line up your new crank trigger to a notch the same time the OEM crank trigger is on a notch. You should have good luck. You will need to adjust the timing by trial and error, but a trip to the drag strip will find the best setting."

Sounds like they only use one edge, the rise or fall, so it should work. I assumed in error, that both edges used. It would be nice to know which edge is used for static timing settings. By capturing the distributor sensor, crank sensor, and coil primary signals on multi channel scope or logic analyzer, will help in your static setup of front sensor.

Standard Hall sensors generate a low signal for metal target viewed by sensor, and a open collector high, with open slot. A pull-up resistor to 5V is supplied by ECM. This will let you easily know what to align.
 
It would be nice to know which edge is used for static timing settings.

I asked that exact question and he told me that if I make the slot narrow enough, I can eyeball it and it should be close enough to run and then adjust it to get the best MPH.

As I recall, when I looked at some motors in the wrecking yard, the pickup was on the trailing edge of the hole at 10 degrees ATDC. But I need to check it again because it could have been BTDC. Either way, seems reasonable to think that the trailing edge is the one to use. Could be wrong, but that is what I am going to start with.

BTW, this was in December of 2009 when I was asking these questions. Been looking at doing this for a very long time, should have had it done way before now. Embarrassing.
 
Thank you for the info about the single edge sensing. The result is similar to what I use for electronic advance ignition where I use a single long tab, and slot, and both edges rise and fall for each cylinder. The max advance and base timing is defined by those. I differ by using an optical vane sensing, for the timing reference.

Hall gear tooth sensors employ an internal AC coupled reference circuit. It is similar to automatic gain control AGC on a radio. Del 67Dart273 will know what I am talking about. It self tunes the threshold so the sensor works over a large temperature range, aging and target variations. It might be why there are pairs of tabs and slots, to double the frequency for more stable signal when cranking. The self tuning improves the sensor position accuracy.

Here is typical information for the application of Hall sensor targets https://www.google.com/url?sa=t&sou...-nL1H4eS8kDH2hdXg&sig2=iBww6maGyqy8FU3FZjOysQ
 
Last edited:
I agree that if you can adjust the sensor location, or rotate the trigger wheel (slotted holes), all that matters is that you are centered in the adjustment range and that the slots are equally spaced. I put a 36-1 trigger wheel on my 1965 273. I bought it off ebay for $40 and the seller custom-made to the OD of my crank pulley. I secured w/ both Loc-tite and set screw (from pulley ID). Looks very similar to wish4hemi's (link in post 6, probably not him since I recall from Oklahoma). I made a bracket to secure either a Ford EDIS VR pickup or Mopar Hall-effect, both off the lower water pump bolts. The Ford EDIS module can decipher the signal, but that sets just base timing. Megasquirt can command spark timing to it. I wonder if Holley's Commander 950 could also since it commands a GM 8-pin HEI module with fairly similar SAW signal. Haven't used yet or even run that engine w/ normal distributor much. BTW, did this years ago, but I don't worry about schedule like DionR (what's a year or 8?), I can just blame the yard, pool, teenager, work, ... A wedding took all last weekend and a funeral this weekend, but those rightly take preference over old Mopars.

36-1 toothed wheel - small.jpg


Ford EDIS sensor - small.jpg


Chrysler sensor - small.jpg
 
Just to update some earlier info, the window in the flexplate starts at about 10 BTDC and ends right around 6 BTDC. This ignores any inaccuracy in the timing marks and was done in a wrecking yard crawling around a rig, so no promises that this info is 100% accurate.
 
I have a 5.2 and 5.9 flex if you want a photo, measurements, etc.
 
The info I have is 11.75" diameter, window .4"x.4" and equal spacing. It would be good if Del verified it though, I didn't take the measurements myself.
 
With 11.75" diameter the circumference is about 36.9". The 10 - 6 = 4 deg window. 36.9 x 4 / 360 = 0.41", sounds very close. The actual electrical signal will be likely less than 4 degrees.
 
I'll get busy

IMG_0909cs1.JPG


IMG_0909cs2.JPG


Don't try to read this, just showing the "functional" as how I did this. Slot to slot leading to leading "around the wheel" most accurate I could manage is bending a machinist's scale. Incidently, I've had this since about 72 in the Navy

This measurement, near as I can see, is 4 and 33/ 64--34/64

IMG_0916cs.JPG


This is what I'm calling the "tangental" measurement slot to slot, leading to leading
4.390

IMG_0915cs.JPG


Diameter/ circumference:

Took circumference First tried to stretch tape around. Then I marked one slot on flat surface, and rolled the "wheel" 1 turn, marked again

Then measured diameter with tape as accurate as I could. I have no large trammel calipers

dia 11 1/2, so pi says circumference is 33.1284

"Rolled" measurement is 36 3/16, so there's some error there (36.1875)

slots measure "about" (they vary) .393--.4. Measured with internal mike, double checked it with external mike

I wish I'd known/ thought about this, I tore 3 engines down, 2 had the distributors "as running." I did punch mark the dist at TDC. I could just as well have mapped the relationship of the trigger wheel sensor.
 
Last edited:
Using 4.39" between leading edges and circumference, yields 43.7 degrees between. If all measure the same, then it possible there are small measurement errors and the true degrees is 45.

A measurement trick I use a 3/8" wide flat measure tape, and start a few inches in to avoid the leader, then subtract it out.
 
Using 4.39" between leading edges and circumference, yields 43.7 degrees between. If all measure the same, then it possible there are small measurement errors and the true degrees is 45.

A measurement trick I use a 3/8" wide flat measure tape, and start a few inches in to avoid the leader, then subtract it out.


Dave that was "tangental" figure, not "around the circle"
 
How does this work? "Around the circle" slot to slot I measured 4 33/64---34/64 or 5.515. That times 8? is is 36.12 which is not too far off of measured dia x pi (33.128)

If we use 4 34/64 that is 4.531, x8 is 36.248. I guess if you can accurately plot 8 positions on a layout by degrees, then all you need is the slot length
 
Slot to slot with caliper gives the "cord", straight line not following the arc. Small distance, very close. The cord can be used, the end points and radius, form a triangle, that bisected is two right triangles, trig solves the angle.

The mechanical, does not follow electrical signal, the sensor head is about 0.08" inside.
 
Ugh, still working on this. I had actually decided to buy an MS3Pro box and dodge the "pay a tuner" bullet, but then I bought a '15 R/T 6M and have decided to go back to the Magnum ECU to save the cost of buying an $1100 box.

So, got my reluctor drawn up with 8 teeth and used the 4 degree notch size, ended up being like .27" wide. Little concerned it might be too small.

Worse yet, found this thread - Megasquirt Support Forum (MSEXTRA) • 426 Gen3 Hemi crate engine No Start (View topic) where the guy had problems with MegaSquirt getting a good crank reading using an '11 and up CKP sensor on an '08 6.1 crank. As it happens, I am using the same '11 sensor on my setup. I'm sure his issue relates to Kit's concerns at the beginning of this thread, but in my mind that doesn't explain why the guy couldn't get a stable reading until he swapped to a matching sensor. I could see him having issues with it if he was running the OEM setup, but MegaSquirt? Maybe it was a bad sensor to begin with and swapping to the earlier sensor more more about a good sensor rather than a different one?

In the end I think I am going to run it the way it is and keep the above sensor issue in the back of my mind. Just looking for any thoughts on the subject before spending money (again - already cut a 36-1 wheel).

Untitled.jpg
 
Different sensors, have different operating points depending on sensor design, target material, shape and proximity. Cylindrical sensors do not trigger at center, they trigger before, and release after. How much before and after depends on sensor, and target related parameters. Professionals that apply sensors for position sensing, typically evaluate sensors, and target in a test setup, then use the imperial evidence for design.

When there is uncertainty about how the ECU handles sensor inputs, sensors signals are typically simulated using specially programmed micro controller. The ECU is then bench tested, monitoring with logic analyzer on sensor signals, and ECU driven outputs tor ignition, fuel and other controls.

I speak from experience, it works. I do not ponder as much, as I get busy test and evaluate, if problems exist the tests indicate what.
 
-
Back
Top